Entropy-Controlled Information Architecture: A Unified Framework for Machine-Optimized Content Delivery and Retrieval

This paper presents Entropy-Controlled Information Architecture (ECIA), a unified framework for optimizing information delivery to autonomous AI agents. Drawing on the Theory of Stupidity (Petrenko, 2025), it formalizes how environmental noise (D) sabotages machine attention (A), causing LLM failures independent of model capability (I). The paper introduces two core implementations: AI Optimization (AIO): A publisher-side protocol for creating machine-centric content layers. Entropy-Controlled Retrieval (ECR): A consumer-side pipeline for noise-free ingestion. Empirical benchmarks demonstrate 100% answer accuracy, 6x faster retrieval, and potential billions in annual savings at search-engine scale.Publication Information: Journal: Science, education, and culture (Наука, образование и культура), 2026. № 1 (75). ISSN 2413-7111.